Author: ParisR

Perplexity Just Put a Salesforce-, HubSpot-, and Snowflake-Connected AI Agent Inside Your Slack — Here’s the SMB GTM Playbook

For most of 2024 and 2025, the limiting reagent in small-business AI wasn’t models — it was integrations. You could ask any chatbot a clever question, but it couldn’t see your Salesforce pipeline, query the HubSpot contacts you actually email, or pull a quarter of revenue from Snowflake. That changed quietly over the last two weeks. Perplexity moved its multi-model agent — branded “Computer” — into the enterprise tier, then on May 4, 2026 shipped an update with a slate of business-grade connectors and stronger model orchestration. Inside one weekend after the enterprise launch, more than 100 enterprise customers reportedly messaged Perplexity demanding access. The reason matters for any SMB go-to-market team: this is the first credibly priced agent that ships pre-wired to the systems your sellers and marketers actually use.

What’s in the box

Perplexity Computer for Enterprise is not a single model — it’s an orchestration layer that routes a query across roughly 20 different AI models and 100+ integrations, picking the right tool for the job. The May 4 update added or hardened business-grade connectors for Snowflake, Datadog, Salesforce, SharePoint, and HubSpot, and lets administrators install custom connectors using Anthropic’s Model Context Protocol (MCP). The single most useful primitive for SMBs is the @computer mention inside Slack: a sales rep, marketer, or founder can drop “@computer pull all Salesforce opportunities over $25K still in ‘qualifying’ as of last quarter, summarize blockers, suggest next-step emails” into a thread, and the agent runs the query, the synthesis, and the draft. The Snowflake and Datadog connectors are exclusive to the enterprise tier, but the ones most SMB GTM teams actually need — Salesforce, HubSpot, SharePoint, Slack — are present.

Why this is a real GTM moment, not just another product launch

Two structural shifts matter here for small-business sales and marketing leaders.

First, Perplexity is explicitly competing with Microsoft and Salesforce by not being a system of record. VentureBeat’s coverage was direct: Perplexity is staking out the orchestration layer as a separate category. For SMBs, that’s a feature, not a bug. You don’t have to rip out HubSpot or Salesforce or your warehouse. You add an agent that talks to all of them in plain English. Coverage from PYMNTS noted Perplexity’s enterprise customers report compressing roughly 3.25 years of work into four weeks on certain workflows. Even if you discount that by 80%, the reframe is real: the bottleneck has moved from “can the tool do this” to “do we know what to ask it.”

Second, the friction to give a non-technical seller or marketer access to the company’s actual data warehouse just collapsed. Historically, asking “what’s our pipeline coverage by segment, weighted by stage, vs. last quarter” required either a data analyst or a fragile Looker dashboard. With Computer for Enterprise, your AE asks the question in Slack and gets a defensible answer. That doesn’t replace your RevOps team; it lets your RevOps team spend their time on the questions that actually require thinking.

A 30-day SMB GTM playbook

If you run sales or marketing at a small business, here is a sequence that turns this from “interesting headline” into measurable pipeline impact.

Week 1 — Pick three questions a human currently answers. Audit one week of your team’s Slack and email. Find three recurring questions that someone answers manually: “what’s our MQL volume this week by source,” “which open opps are stalled and what was the last touch,” “what content actually converted last month.” Those become your first three @computer prompts.

Week 2 — Wire two connectors, not five. Resist the temptation to connect everything. Pick the two systems that contain 80% of the answers — usually your CRM (HubSpot or Salesforce) plus your communication system (Slack or email). Get those clean and authenticated. Validate that @computer answers your three baseline questions correctly.

Week 3 — Hand it to one rep and one marketer. Pilot with two people, not the whole team. Have them each replace one weekly manual task with @computer. Time the difference. Capture before-and-after metrics. This is the dataset that buys you internal credibility for a broader rollout.

Week 4 — Build one prompt library and one guardrail. Codify the prompts that worked into a shared doc. Add one guardrail rule (e.g. “no agent-written external email goes out without human edit”). That is the difference between “we tried Perplexity Computer once” and “this is now part of how we sell.”

If you’d rather not assemble that playbook from scratch, LevelUpLabs.co is a membership purpose-built for entrepreneurs and SMB GTM leads who want ready-to-deploy AI workflows. It includes a prompt library tuned for sales and marketing, video walkthroughs of agent-driven GTM stacks, checklists you can hand to a junior team member, and partner discounts on the tools you’d otherwise pay retail for. It’s the operator-level companion to news like this — built so you can ship in days instead of quarters.

The takeaway

The headline isn’t that Perplexity launched another agent. It’s that an AI agent with native Salesforce, HubSpot, Snowflake, SharePoint, and Slack access is now a button-click away for any small business willing to set it up — and the early enterprise demand suggests the early adopters are already moving. The SMB GTM teams that quietly ship the first three @computer workflows in May will look very different from the ones still debating whether to “pilot” by Q3.

Sources:

You Can’t Skip Classic SEO to Win AI Search — The 92% Correlation Nobody Wants to Hear

A founder asked me last week if he could “skip the SEO stuff” and just optimize for ChatGPT and Perplexity. He’d read three breathless posts about GEO, AEO, and citation engineering. He thought traditional SEO was the past.

Here’s what I told him, and what every recent dataset confirms: there is roughly a 92% correlation between pages that rank in the organic top 10 on Google and pages that get cited in Google AI Overviews. Your blue-link rankings are not a separate world from AI visibility. They are the on-ramp.

That is the unglamorous truth nobody wants to publish in a hot take.

The mechanic — why retrieval looks a lot like ranking

AI search engines don’t read the open web in real time. They use retrieval pipelines: candidate pools, relevance scoring, freshness filters, source-quality weighting. The candidate pool for ChatGPT, Perplexity, AI Overviews, and Gemini is built on the same signals classical search uses — crawlability, internal linking, topical depth, query-page relevance, authority.

If a page can’t be crawled cleanly, an LLM can’t ingest it. If it isn’t relevant for the query, it isn’t pulled into the candidate set. If it’s slow, the embedding pipeline deprioritizes it (pages with FCP under 0.4s average 6.7 citations versus 2.1 for over 1.13s). Bad SEO fundamentals are upstream of bad AI visibility. No amount of “GEO content” survives a broken crawl path.

Then the AI layer adds its own filters on top — answer-unit structure, statistic and quote density, entity consistency, fresh updates. But those filters operate on the candidate pool that classical SEO built. If you’re not in the pool, none of it matters.

What a “skip the basics” strategy actually looks like

I’ve audited a dozen sites this year that bought the “GEO is different” pitch. Same pattern every time: thin technical SEO, slow page speed, broken internal links, no entity consistency between site and Wikidata, no schema, and a flurry of new “ChatGPT-optimized” content sitting in folders Googlebot has never visited. Their AI visibility was zero. Of course it was — they were invisible to the layer underneath.

The fix is never spectacular. It’s the same boring list it’s been for fifteen years, with a few modern items added at the end:

  • Crawlable site, clean canonical tags, no orphan pages
  • Strict H1 → H2 → H3 hierarchy on every important page (68.7% of cited pages do this)
  • Internal links from authoritative pages to deep ones
  • Page speed under 1 second to FCP wherever possible
  • Real backlinks from real sites — sites with 32K+ referring domains are 3.5× more likely to be cited by ChatGPT than sites with under 200
  • Then layer on the new stuff — entity work, citation-engineered paragraphs, freshness cadence, schema

The new stuff is real. It just sits on top of the old stuff. Skip the foundation and you are decorating a building you haven’t poured.

Why the 92% number matters strategically

A 92% overlap between the organic top 10 and AI Overview citations means something specific: ranking work is dual-purpose. Every dollar spent making a page rank #6 is also a dollar spent making it citable. You don’t have to fund two parallel programs. You have to fund one program that ends in two outcomes.

That reframes the budget conversation in every agency-client call I’ve had this year. “GEO” isn’t a separate line item. It’s a layer added to the SEO line item. Anyone selling it as a parallel discipline is either confused or selling 2× the hours for 1× the work.

What to do this week

1. Pull your top-10 ranking list. Filter for any page on a high-intent commercial query. That is your AI-citation candidate pool. Fix indexability, page speed, internal links, and heading hierarchy on those pages first.

2. Compare your ranking pages to your AI-cited pages. Run a manual prompt audit — ask ChatGPT, Perplexity, AI Overviews, and Gemini your top 20 queries and log which URLs they cite. The overlap tells you where retrieval is doing its job. The non-overlap tells you which pages need on-page AI work — answer units, stats, quotes, entity reinforcement.

3. Audit any “GEO-only” content for traditional SEO basics. If a page can’t be reached in three internal clicks from your homepage, it isn’t in any candidate pool — AI or otherwise.

4. Stop writing new “GEO content” until the foundational pages are clean. New articles inherit the site’s authority signals. Pouring content onto a broken site is the slowest possible way to be cited.

The agencies pitching “GEO replaces SEO” are selling a story. Retrieval pipelines, candidate pools, and a 92% top-10 correlation are saying something quieter and more useful: do the unsexy work, then add the new layer.

If you’re a brand that wants to be the answer LLMs reach for (not just rank on Google), Paris Roussos has been engineering search visibility for 30 years and now runs done-for-you AI SEO. Flat-rate, no-fuss. Email parisroussos@gmail.com.

The future of AI SEO looks a lot like the past — only the people who kept doing the boring parts are getting cited.

Only 7% of Small Businesses Are Running Production AI Agents — Here’s the GTM Playbook to Be in That 7% Before Q3

If you only read one number from the spring 2026 marketing reports, make it this one: 7%. That’s the share of SMBs running production AI agents, according to OneReach.ai’s 2026 agentic AI benchmarks. The same study clocks enterprise teams at 34% and mid-market at 19%. Salesforce’s State of Marketing 2026 puts the global “marketers using agentic AI today” number at just 13%, even though 80% of SMB marketers expect major-to-moderate ROI improvements from AI agents. Translation: the technology has clearly arrived, the bigger players are already deploying, and most small-business GTM teams are sitting in the meeting where someone says “we should look into this.”

The window to be early is narrow. BizBuySell’s Q1 2026 Insight Report shows 63% of small businesses now use AI broadly, with 83% reporting performance gains. ChatGPT leads adoption at 82%, followed by Gemini (50%), Claude (39%), Copilot (25%), and Grok (18%). What that actually means is: most small businesses now use AI for one-off tasks, but only a tiny minority have promoted any of it to a running, scheduled, measurable agent inside their go-to-market motion. That’s the gap. And it’s the cheapest one to close before the rest of the market catches on.

What “production agent” actually means in a GTM context

A production agent isn’t a chat window you open when you remember to. It runs without you. In SMB go-to-market, the four highest-leverage candidates right now are:

1. Inbound qualification. An agent that watches every form fill, free-trial signup, and meeting request, classifies fit (ICP / not-ICP), enriches the company data, drafts the first reply, and either books the meeting directly or routes the lead to you with a one-paragraph brief. Replaces the “I’ll triage these tomorrow” pile that quietly costs you 15–25% of warm leads every month.

2. Outbound personalization. An agent that takes a target list, researches each company’s recent news, public hiring patterns, and product changes, and writes a first-touch email tied to a real signal. Stops the “Hi {{first_name}}” copy-paste that everyone now ignores.

3. Meeting prep + follow-up. An agent that reads the prospect’s site, your CRM history, and the calendar invite, then drops a pre-call brief into Slack 30 minutes before the meeting and a structured follow-up + next-step email within 10 minutes after.

4. Content repurposing. A scheduled agent that takes your one piece of long-form content per week (podcast, webinar, article) and ships the LinkedIn post, the X thread, the email newsletter section, and three short-form video scripts. Salesforce’s data shows this kind of agent frees 6–7 hours per week per marketer.

Pick one to start. Not four. The most common reason SMB teams stall on agentic AI is choosing too many candidates and shipping zero. The 7% number above is small precisely because most teams attempt all four at once and none of them ever go live.

The 30-day SMB GTM playbook

Week 1: pick the single workflow above where you already feel the pain. Document the current manual version step by step. This document is your spec.

Week 2: build it on a stack you already pay for — HubSpot Breeze Agents, Salesforce Agentforce (now actually priced for SMBs), Klaviyo Flows AI for ecom, or a no-code orchestrator like n8n / Zapier piped into Claude or ChatGPT API. Don’t switch CRMs to do this. Switching CRMs is how this dies.

Week 3: ship it on 20% of traffic or 20% of leads. Track exactly two numbers: time-to-first-touch and conversion at the next stage. Compare to your manual baseline.

Week 4: keep it, kill it, or fix it. If it’s better, scale to 100%. If it’s worse, fix the prompt or the data and try again. If it’s wildly worse, kill it and pick the next workflow.

This is the part of the cycle where the membership at LevelUpLabs.co earns its keep for entrepreneurs and small-business GTM operators. Instead of building the prompt, the workflow doc, and the measurement spreadsheet from scratch, you get prompt libraries already wired for inbound qualification and outbound personalization, video walkthroughs of the agents working end-to-end, ready-to-use checklists for the 30-day deployment cycle above, and partner discounts on the tools you’d otherwise stack at full price. It’s the difference between watching the 7% and joining it.

The bottom line

The Salesforce report is not telling SMBs they’re behind — it’s telling them they’re early. 80% of SMB marketers expect major ROI improvements from agentic AI. Only 7% have actually shipped one. The arithmetic on competitive advantage is not subtle: pick one workflow, ship one agent in 30 days, measure the two numbers, and you’ll be ahead of 93% of small-business competitors before this calendar quarter ends.

Sources:

Your Buyers Are About to Treat AI Agents as Headcount — Here’s Why Per-Seat Pricing Is Quietly Dying

Your Buyers Are About to Treat AI Agents as Headcount — Here’s Why Per-Seat Pricing Is Quietly Dying

For two decades, B2B SaaS pricing has rested on one assumption: the buyer is paying for human seats. License tiers, usage caps, contract negotiation, expansion math, and customer-success comp plans were all built around this. In 2026, that assumption is breaking — and the GTM teams that haven’t noticed are about to walk into renewal cycles where the buyer’s procurement question has changed entirely.

The data point that should set off every revenue leader’s alarm: nearly nine out of ten manufacturing leaders say they expect to use AI agents as additional workforce capacity within the next 12 to 18 months, according to the World Economic Forum’s 2026 Future of Jobs reporting. Gartner’s parallel forecast — that 40% of enterprise applications will embed AI agents by the end of 2026, up from less than 5% in 2025 — describes the supply side of the same shift. Buyers are explicitly modeling agents as units of work, not as features. PwC’s 2026 AI Business Predictions reinforce the framing: workers with AI skills are commanding wage premiums up to 56%, but the more interesting line in the report is that organizations are increasingly translating “AI capability” into FTE-equivalent capacity inside their workforce planning models. When your buyer says “we need 30% more output next year,” the answer is no longer “hire 30% more people” — it’s a blended question that includes agents.

What that means in your sales motion is concrete. Procurement is starting to ask three new questions on RFPs that didn’t appear in 2024: how does your tool meter agent activity (since agents will be the heaviest users), how do you price when the same workflow is run by a human one quarter and an agent the next, and what’s your outcome-based pricing option. Per-seat SaaS pricing breaks under all three. If a buyer pays per seat and then deploys an agent that does the work of three seats, who’s the seat? If your contract caps usage by named user, an agent setup with one service account looks like one seat — but consumes the resources of fifteen. And if the buyer’s CFO is modeling spend against work-completed rather than against headcount, your pricing model is on the wrong side of the unit economics conversation.

The GTM rewrite is not optional, and it’s not just a finance issue. Three things change in the same renewal cycle. First, the discovery question shifts from “how many users will need access” to “what workflows will agents run, and at what frequency.” That’s a different conversation, and it favors the seller who shows up with usage telemetry and a per-task cost model rather than a per-seat menu. Second, the RFP response template has to include an outcome- or task-based pricing option even if your default is still per-seat — because a meaningful number of buyers will only shortlist vendors who offer it. Third, customer success comp needs a new metric: not “seats activated” but “agent runs completed against the contracted workflow.” Sellers and CSMs need shared definitions or expansion will leak.

If you want a steady feed of signals like this — curated trend reporting written for CEOs and founders, not data scientists — bookmark TrendInsightsJournal.com. It’s where these moves get tracked weekly so you can spot the meaningful shifts (AI, crypto, macro, metatrends) without drowning in feed noise. Read the brief, run your week.

There’s a second-order GTM effect worth flagging: when buyers start treating agents as workforce, they also start expecting vendor enablement to look like onboarding for that workforce. Salesforce’s 2026 agent research notes that buyer-side agent deployment is increasingly trusted to make decisions within well-defined boundaries — which means your buyer’s agent will, at some point, be reading your docs, running your demo recordings, and parsing your pricing page on the buyer’s behalf. If your content is human-readable but agent-hostile (PDFs without text layers, sales decks behind gated forms, pricing as a “contact us” wall), you’re invisible to a growing fraction of the buyer’s evaluation surface. Revenue marketing now has a structured-data problem.

The four-part fix for May: add a per-task or outcome-based line item to your pricing menu (even as a “talk to us” option), update your RFP boilerplate with agent-meter language, brief your CSMs on the workflow-completed metric, and audit your top ten pieces of mid-funnel content for machine-readability. The vendors who do this in Q2 will renew at higher unit economics than the ones still selling seats in 2027.

Sources: World Economic Forum (Future of Jobs 2026; how AI will affect work in different industries), PwC (2026 AI Business Predictions; Global AI Jobs Barometer), Gartner (40% enterprise application embed forecast), Salesforce (8 Ways AI Agents Are Evolving in 2026), Harvard Business Review (9 Trends Shaping Work in 2026 and Beyond).

TikTok Just Handed Small Businesses the “Override” Button on Smart+ — Here’s the Q2 Playbook

For two years, TikTok’s Smart+ has been the platform’s answer to Meta’s Advantage+ — a black box that promises better ROAS in exchange for handing over the controls. SMB advertisers liked the conversion lift. They did not like the part where they couldn’t see what the AI was actually doing. In TikTok’s Q2 2026 Product Preview — published in early May — that complaint finally got a fix, and it changes the small-business GTM playbook for the rest of the year.

What changed in Smart+

The headline update is module-level controls. Instead of “Smart+ on” or “Smart+ off” as a single switch, advertisers can now turn AI automation on or off for each piece of a campaign — targeting, budget, and placements — independently. Every module shows a Smart+ label so you can tell at a glance what’s automated and what’s manual. Smart+ has also been built directly into Traffic campaign setup, which means link-click campaigns (the bread and butter of small business lead-gen) get the same modular toggle that previously only existed on Smart+ Web Campaigns and Catalog Ads. TikTok confirmed all of these updates will roll out globally in Q2 2026.

Two other pieces of the Q2 preview matter for SMBs specifically. TikTok Pulse — the brand-safety placements that put your ad next to the top 4% of TikTok content — got a performance-focused refresh, making it a more credible buy for direct-response advertisers, not just brand-budget spenders. And on the commerce side, TikTok Shop’s Product Amplification Program (originally launched February 2026) keeps expanding, designed to reduce setup friction and surface inventory inside organic content.

Why this matters for small businesses doing GTM on TikTok

Smart+ has been delivering real efficiency — but for SMBs, full-automation campaigns have one persistent failure mode: the AI optimizes toward whatever conversion event it can find, even when that’s not the conversion event your business actually cares about. A local services business gets clicks from kids in a different metro. A DTC product company spends 80% of its budget on a single placement that converts cheap but doesn’t retain. The fix used to be “turn Smart+ off and run manual” — which gave up the lift entirely.

Module-level controls let SMB advertisers do something more surgical. The most common winning configuration on Meta’s equivalent system has been: keep AI in charge of creative testing and bidding, but lock down placement, geography, and audience guardrails yourself. TikTok now supports the same posture. For a small business, that’s the difference between “Smart+ printed money” and “Smart+ printed money for the wrong customer.”

The numbers underneath this matter, too. eMarketer’s coverage of Smart+ adoption shows performance lifts in the 30–50% range on conversion campaigns when properly configured — but only when the underlying data is clean and the campaign objective matches the business’s actual revenue event. The new modular toggles only help if you’ve done the homework on conversion mapping first.

A 7-day SMB playbook to get ahead of competitors

Most small businesses won’t touch their TikTok account in the next two weeks. Here’s how to use that gap.

Day 1–2: Audit your existing Smart+ campaigns. For each one, identify the single module where you have the strongest first-party data — usually geography or first-party audience. That’s the module you’ll turn manual.

Day 3–4: Tighten your conversion event. If you’re still optimizing toward “page view” or “add to cart,” fix it. The new Smart+ controls are only as smart as the goal you give them. Lock the optimization event to the actual revenue moment — purchase, qualified lead form, booked appointment.

Day 5: Spin up a Smart+-on-Traffic campaign. Now that Smart+ lives inside Traffic, the cost-per-click ceiling for top-of-funnel work drops materially. Use it for warm-audience retargeting first; that’s where the lift is most reliable.

Day 6–7: Test TikTok Pulse for direct response. Pulse used to be a brand spend; the Q2 update makes it credible for performance. Run a small test against your current best-performing placement and watch CTR + post-click quality, not just CPM.

If you’d rather not stitch this together solo, LevelUpLabs.co is built for exactly this kind of moment — a place where entrepreneurs trade live AI playbooks, prompt libraries, video training, ready-made checklists, and partner discounts on the tools you’d actually run a Smart+ test with. It’s the difference between reading another marketing thinkpiece and shipping a campaign by Friday.

The takeaway for SMB marketers

TikTok’s Q2 update isn’t a feature drop. It’s an admission that “fully automated AI ads” was always going to lose to “AI plus the operator who knows their customer.” Small businesses that exploit that — by deciding which parts of the campaign to hand over and which to keep — will outperform whichever competitor is still running blind Smart+ for another quarter. The window where this is a competitive advantage is small. Take it.

Sources:

Google Just Replaced “Call Duration” With AI Lead Scoring in Google Ads — Here’s the SMB Playbook to Win Before Your Competitors Notice

If you spend money on Google Ads and your business takes phone calls — local services, healthcare-adjacent (where allowed), home services, automotive, B2B with a phone CTA — Google quietly changed the rules of your campaigns on April 21, 2026. The change is AI-Qualified Call Conversions, and most SMB advertisers haven’t even seen it yet because it doesn’t require a UI change. It just rewires what a “conversion” means under the hood.

What changed, in plain English

Until April 2026, Google Ads decided whether a phone call from one of your ads counted as a conversion based mostly on call duration — typically 60+ seconds. That worked for about a decade. It also broke for about a decade: spam calls, wrong numbers, and “do you have hours” calls all crossed the duration threshold, while a 45-second call that ended in “I’d like to book” did not.

The new system uses AI to listen to the call recording, extract intent, and decide whether the conversation reflects genuine purchase interest before it’s counted as a conversion. Call duration becomes a secondary signal, used only when recording isn’t available. Advertisers also get AI-generated call summaries and intent tags inside Google Ads — so you can see, per call, what the AI thought happened.

The feature is currently U.S./Canada only. Call recording is on by default for most advertisers (with industries like healthcare and financial services excluded). It was first reported by Hana Kobzova and PPC News Feed.

Why this is a bigger SMB story than it sounds

Three reasons this matters more than the average Google Ads update:

1. Smart Bidding now optimizes against actual quality. Google’s bid algorithms have always optimized toward whatever you tell them is a conversion. When the conversion signal becomes “calls that show real intent” instead of “calls that lasted >60 seconds,” the algorithm starts buying better leads, not just more leads. The advertisers whose accounts are configured correctly will see CPL drop and pipeline quality rise — without changing a single bid manually.

2. Reporting transparency closes the trust gap. Every SMB owner who has ever called a Google Ads rep to argue about a phantom conversion now has a per-call summary they can read. “Caller asked about pricing for a furnace install, requested a callback Thursday” is a different conversation than the one where you have to take the rep’s word for it.

3. The competitive window is narrow. Google didn’t roll this out with a press tour. Many agencies and in-house teams are running campaigns right now without realizing the conversion definition has shifted underneath them. The SMBs whose advertisers tune to the new signal first will compound that advantage for the rest of the quarter.

The SMB GTM playbook for the next 30 days

Concrete moves that turn this update into pipeline:

  • Audit your call recording and conversion settings today. In Google Ads → Tools → Conversions, confirm call recording is enabled and AI-Qualified Call Conversions is being applied to your call-tracking conversion actions. If you previously turned off recording, turn it on (where compliant) — without it, you fall back to the old duration-based system and lose the upside.
  • Re-baseline before you change bids. Pull a 30-day report on call conversions, average CPL, and pipeline-from-calls before the new signal fully reshapes the data. You want to know your “before” so you can prove the lift to your boss, your client, or yourself.
  • Read the AI call summaries weekly. They are a free goldmine of objection patterns and qualification gaps. If 30% of your calls are people asking about a service you don’t offer, your ad copy is broken — fix it.
  • Tighten your phone-call landing pages. AI-Qualified now penalizes ad clicks that produce off-intent calls. Pages that pre-qualify (price ranges, service area, hours, “we don’t do X”) will see CPL improve. Pages that bait every click will see CPL get worse.
  • For multi-location SMBs, route by intent tag. Once Google exposes intent tags consistently, you can route high-intent calls to your best closer and informational calls to a generalist or chatbot. That’s a pipeline-velocity win, not just an ad win.

Where the agency conversation breaks (and how to win it)

If you have an outside agency running your Google Ads, this is the right week to ask exactly two questions: “Are AI-Qualified Call Conversions enabled on our account, and what was our duration-based conversion rate vs. the new AI-qualified rate over the last 14 days?” The quality of the answer is the quality of the agency. Good ones will have already adjusted; mediocre ones will not have noticed.

This is exactly the kind of edge LevelUpLabs.co is built to give entrepreneurs — a place to skip the “is this real?” wondering and get straight to the playbook. Inside the membership: AI-prompt libraries for ad copy and landing pages, video training that walks through tactical settings like the ones above, ready-to-use checklists, and partner discounts on the tools you’d otherwise be evaluating one by one. It’s the difference between reading about an update and shipping a campaign change because of it.

The takeaway

The phone-call conversion is one of the oldest pieces of the small-business GTM stack — and Google just rebuilt it on AI. The advertisers who treat April 21, 2026 as the day their conversion definition changed (and adjust accordingly) will be quietly compounding lower CPL and better pipeline through the rest of Q2. The advertisers who don’t read this update will keep wondering why their ROAS feels off.

Sources:

test test